Waymo Rolls Out Its Sharpest-Eyed Robotaxi Yet, the 6th-Generation Driver

Waymo has just enabled fully driverless trips with their sixth-generation Driver. Employees and guests are now getting into sleek Ojai minivans and driving through Phoenix streets with no one behind the wheel. Public access is just around the corner, and the company hopes to see a significant increase, up to 1 million paid trips per week, by the end of the year.

The hardware stands out for being able to accomplish a great deal with far less resources. Previous configurations required 29 cameras, but this one only need 13: four lidar devices replace the five earlier ones, and six radar units handle the remainder. Total sensors are decreased by 42%, allowing the system to see further, clearer, and even in harsher situations. Custom silicon chips that handle intensive processing onboard reduce components and prices while keeping everything functioning smoothly. The entire package costs less than half of the 5th-generation system, making it much easier for the Metro Phoenix factory to produce tens of thousands of vehicles each year.

Cameras are at the heart of the vision system, with each holding a 17-megapixel imager that captures millions of data points per frame. The resolution is far superior to that of typical automobile cameras, allowing them to capture fine details in deep shadows, strong glare, and dim streetlights. Another thing they do well is keep trucking even during heatwaves or frigid temperatures. You need fewer cameras to complete the task because each one collects more useful information. They include built-in cleaning mechanisms to remove rain, filth, and ice, ensuring that the vistas remain pristine. When fog or severe rains reduce visibility, LiDAR and radar come in to fill the gaps.


LiDAR uses laser pulses to create these extremely detailed 3D maps. Engineers added specialized optics and processors, allowing it to reach further and cut through adverse weather that would otherwise scatter communications. The short-range devices provide pinpoint accuracy surrounding the vehicle, making them ideal for detecting pedestrians approaching too near or determining door clearance in tight spaces. It also easily handles distortion caused by flashy signs or road spray on freeways.


Even when cameras fail to monitor objects by distance, speed, and size, radar does. These super-dense, real-time maps are created using new chipsets and proprietary algorithms and can withstand extreme weather conditions. Lightweight machine-learned models collect all of the data from all of the sensors and allow each one to work independently without interfering with the others. They also have external audio receivers that detect sirens, train horns, and crossing bells and can locate their location even before you see them. High-speed wind noise is filtered out, allowing the system to react quickly to emergency vehicles.